Setting up environment¶

In [1]:
import sys
import os
sys.path.insert(0, os.path.abspath("../"))
sys.path.insert(0, os.path.abspath("../nsforest/nsforesting"))
import numpy as np
import pandas as pd
import scanpy as sc
import matplotlib.pyplot as plt
import nsforest as ns
from nsforest import utils
from nsforest import preprocessing as pp
from nsforest import nsforesting
from nsforest import evaluating
from nsforest import plotting as pl
from IPython.display import Image

Data Exploration¶

In [2]:
# Loading h5ad
data_folder = "../demo_data/"
file = data_folder + "adata_layer1.h5ad"
adata = sc.read_h5ad(file)
adata
Out[2]:
AnnData object with n_obs × n_vars = 871 × 16497
    obs: 'cluster'
In [3]:
cluster_header = "cluster"
In [4]:
# Looking at sample labels (obs_names)
adata.obs_names
Out[4]:
Index(['A01_1_Nuclei_NeuNP_H200_1025_MTG_layer1_BCH9',
       'A01_BCH3_1NeuNP_H200.1030_MTG_Layer_1',
       'A02_BCH1_1NeuNP_H200.1025_MTG_layer_1',
       'A03_1_Nuclei_NeuNP_H200_1025_MTG_layer1_BCH9',
       'A04_1_Nuclei_NeuNP_H200_1025_MTG_layer1_BCH9',
       'A04_BCH1_1NeuNP_H200.1025_MTG_layer_1',
       'A04_BCH3_1NeuNP_H200.1030_MTG_Layer_1',
       'A05_1_Nuclei_NeuNP_H200_1025_MTG_layer1_BCH9',
       'A05_BCH1_1NeuNP_H200.1025_MTG_layer_1',
       'A05_BCH3_1NeuNP_H200.1030_MTG_Layer_1',
       ...
       'P09_1_Nuclei_NeuNN_H200_1025_MTG_layer1_BCH7',
       'P09_1_Nuclei_NeuNN_H200_1025_MTG_layer1_BCH9',
       'P09_1_Nuclei_NeuNN_H200_1030_MTG_layer1_BCH8',
       'P09_BCH1_1NeuNN_H200.1025_MTG_layer_1',
       'P10_1_Nuclei_NeuNN_H200_1025_MTG_layer1_BCH6',
       'P10_1_Nuclei_NeuNN_H200_1025_MTG_layer1_BCH9',
       'P10_BCH1_1NeuNN_H200.1025_MTG_layer_1',
       'P11_1_Nuclei_NeuNN_H200_1025_MTG_layer1_BCH7',
       'P11_1_Nuclei_NeuNN_H200_1025_MTG_layer1_BCH9',
       'P11_1_Nuclei_NeuNN_H200_1030_MTG_layer1_BCH8'],
      dtype='object', length=871)
In [5]:
# Looking at genes (var_names)
# Note: you can change the var.index to be ensembl id or gene name
# adata.var_names = adata.var["feature_name"]
adata.var_names
Out[5]:
Index(['A1CF', 'A2M', 'A2M_AS1', 'A2ML1', 'A2ML1_AS1', 'A2MP1', 'A3GALT2',
       'A4GALT', 'AAAS', 'AACS',
       ...
       'ZUFSP', 'ZW10', 'ZWILCH', 'ZWINT', 'ZXDC', 'ZYG11A', 'ZYG11B', 'ZYX',
       'ZZEF1', 'ZZZ3'],
      dtype='object', length=16497)
In [6]:
# Checking cell annotation sizes 
adata.obs[cluster_header].value_counts()
Out[6]:
e1_e299_SLC17A7_L5b_Cdh13     299
i1_i90_COL5A2_Ndnf_Car4        90
i2_i77_LHX6_Sst_Cbln4          77
i3_i56_BAGE2_Ndnf_Cxcl14       56
i4_i54_MC4R_Ndnf_Cxcl14        54
g1_g48_GLI3_Astro_Gja1         48
i5_i47_TRPC3_Ndnf_Car4         47
i6_i44_GPR149_Vip_Mybpc1       44
i7_i31_CLMP_Ndnf_Cxcl14        31
g2_g27_APBB1IP_Micro_Ctss      27
i8_i27_SNCG_Vip_Mybpc1         27
i9_i22_TAC3_Vip_Mybpc1         22
g3_g18_GPNMB_OPC_Pdgfra        18
i10_i16_TSPAN12_Vip_Mybpc1     16
g4_g9_MOG_Oligo_Opalin          9
i11_i6_EGF_Vip_Mybpc1           6
Name: cluster, dtype: int64

Preprocessing¶

In [7]:
cluster_header = "cluster"
medians_header = "medians_" + cluster_header
binary_scores_header = "binary_scores_" + cluster_header
In [8]:
# Running scanpy dendrogram
ns.pp.dendrogram(adata, cluster_header, save = "_" + cluster_header + ".png") 

WARNING: dendrogram data not found (using key=dendrogram_cluster). Running `sc.tl.dendrogram` with default parameters. For fine tuning it is recommended to run `sc.tl.dendrogram` independently.
WARNING: You’re trying to run this on 16497 dimensions of `.X`, if you really want this, set `use_rep='X'`.
         Falling back to preprocessing with `sc.pp.pca` and default params.

C:\Users\bpeng\AppData\Local\anaconda3\Lib\site-packages\anndata\_core\anndata.py:522: FutureWarning: The dtype argument is deprecated and will be removed in late 2024.
  warnings.warn(

WARNING: saving figure to file figures\dendrogram_cluster.png
Out[8]:
AnnData object with n_obs × n_vars = 871 × 16497
    obs: 'cluster'
    uns: 'dendrogram_cluster'
    obsm: 'X_pca'
In [9]:
# Calculating cluster medians per gene
ns.pp.prep_medians(adata, cluster_header, use_mean = False, positive_genes_only = True) 

Calculating medians...

Calculating medians (means) per cluster: 100%|█████████████████████████████████████████| 16/16 [00:00<00:00, 55.69it/s]

Saving calculated medians as adata.varm.medians_cluster
--- 0.29226064682006836 seconds ---
Only positive genes selected. 11688 positive genes out of 16497 total genes
Out[9]:
AnnData object with n_obs × n_vars = 871 × 11688
    obs: 'cluster'
    uns: 'dendrogram_cluster'
    obsm: 'X_pca'
    varm: 'medians_cluster'
In [10]:
# Calculating binary scores per gene per cluster
ns.pp.prep_binary_scores(adata, cluster_header, medians_header)
adata

Calculating binary scores...

Calculating binary scores per cluster: 100%|███████████████████████████████████████████| 16/16 [01:18<00:00,  4.90s/it]

Saving calculated binary scores as adata.varm.binary_scores_cluster
--- 78.71559691429138 seconds ---
median: 0.0
mean: 0.1428629848824932
std: 0.23137174460615756
Out[10]:
AnnData object with n_obs × n_vars = 871 × 16497
    obs: 'cluster'
    uns: 'dendrogram_cluster'
    obsm: 'X_pca'
    varm: 'medians_cluster', 'binary_scores_cluster'
In [11]:
a = plt.hist(adata.varm[medians_header].unstack(), bins = 100)
a = plt.title(f'{file.split("/")[-1].replace(".h5ad", "")}: {medians_header} histogram')
a = plt.xlabel(medians_header)
a = plt.yscale("log")
plt.show()
In [12]:
a = plt.hist(adata.varm[binary_scores_header].unstack(), bins = 100)
a = plt.title(f'{file.split("/")[-1].replace(".h5ad", "")}: {binary_scores_header} histogram')
a = plt.xlabel(binary_scores_header)
a = plt.yscale("log")
plt.show()
In [13]:
# Saving as new h5ad
filename = file.replace(".h5ad", "_preprocessed.h5ad")
print(f"Saving new anndata object as...\n{filename}")
adata.write_h5ad(filename)

Saving new anndata object as...
../demo_data/adata_layer1_preprocessed.h5ad
In [14]:
outputfilename_prefix = cluster_header
output_folder = "../outputs_layer1/"

Running NSForest¶

Note: Skip if input marker list csv is included¶
In [15]:
# Adding this line to actually subset to positive genes if running NSForest
adata = ns.pp.prep_medians(adata, cluster_header, use_mean = True, positive_genes_only = True) 
results = nsforesting.NSForest(adata, cluster_header, medians_header, binary_scores_header, 
                               output_folder = output_folder, outputfilename_prefix = outputfilename_prefix)

Calculating medians...
use_mean is True. Using the mean expression of each gene per cluster. 

Calculating medians (means) per cluster: 100%|████████████████████████████████████████| 16/16 [00:00<00:00, 102.18it/s]
Saving calculated medians as adata.varm.medians_cluster
--- 0.16356277465820312 seconds ---
Only positive genes selected. 16419 positive genes out of 16497 total genes










    
    






Preparing adata...
--- 0.002991914749145508 seconds ---
Pre-selecting genes based on binary scores...
	 Threshold (mean + 2 * std): 0.6069627661813182
	Average number of genes after gene_selection in each cluster: 1077.125
Saving number of genes selected per cluster as...
../outputs_layer1/cluster_gene_selection.csv
Number of clusters to evaluate: 16










    
    






running NSForest on all clusters:   0%|                                                         | 0/16 [00:00<?, ?it/s]









    
    






1 out of 16:
	e1_e299_SLC17A7_L5b_Cdh13
	 Pre-selected 2029 genes to feed into Random Forest.










    
    






running NSForest on all clusters:   6%|███                                              | 1/16 [00:07<01:53,  7.59s/it]









    
    






	['TESPA1', 'SATB2']
	fbeta: 0.9850034083162916
2 out of 16:
	g1_g48_GLI3_Astro_Gja1
	 Pre-selected 727 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  12%|██████▏                                          | 2/16 [00:10<01:10,  5.04s/it]









    
    






	['LINC00498']
	fbeta: 0.9500000000000001
3 out of 16:
	g2_g27_APBB1IP_Micro_Ctss
	 Pre-selected 485 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  19%|█████████▏                                       | 3/16 [00:13<00:52,  4.04s/it]









    
    






	['CSF2RA', 'ADAM28']
	fbeta: 0.9756097560975611
4 out of 16:
	g3_g18_GPNMB_OPC_Pdgfra
	 Pre-selected 453 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  25%|████████████▎                                    | 4/16 [00:16<00:43,  3.63s/it]









    
    






	['GPNMB', 'OLIG2']
	fbeta: 0.8620689655172413
5 out of 16:
	g4_g9_MOG_Oligo_Opalin
	 Pre-selected 693 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  31%|███████████████▎                                 | 5/16 [00:20<00:38,  3.53s/it]









    
    






	['ST18']
	fbeta: 1.0
6 out of 16:
	i10_i16_TSPAN12_Vip_Mybpc1
	 Pre-selected 1445 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  38%|██████████████████▍                              | 6/16 [00:24<00:38,  3.89s/it]









    
    






	['TSPAN12', 'CHRNB3']
	fbeta: 0.8035714285714286
7 out of 16:
	i11_i6_EGF_Vip_Mybpc1
	 Pre-selected 2393 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  44%|█████████████████████▍                           | 7/16 [00:28<00:34,  3.88s/it]









    
    






	['EGF', 'NRG3_AS1']
	fbeta: 0.7142857142857143
8 out of 16:
	i1_i90_COL5A2_Ndnf_Car4
	 Pre-selected 477 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  50%|████████████████████████▌                        | 8/16 [00:32<00:31,  3.89s/it]









    
    






	['COL5A2', 'FREM1']
	fbeta: 0.8870967741935484
9 out of 16:
	i2_i77_LHX6_Sst_Cbln4
	 Pre-selected 520 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  56%|███████████████████████████▌                     | 9/16 [00:37<00:28,  4.13s/it]









    
    






	['LHX6']
	fbeta: 0.8166189111747851
10 out of 16:
	i3_i56_BAGE2_Ndnf_Cxcl14
	 Pre-selected 308 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  62%|██████████████████████████████                  | 10/16 [00:40<00:24,  4.06s/it]









    
    






	['BAGE2', 'SYT10']
	fbeta: 0.78125
11 out of 16:
	i4_i54_MC4R_Ndnf_Cxcl14
	 Pre-selected 408 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  69%|█████████████████████████████████               | 11/16 [00:44<00:19,  3.89s/it]









    
    






	['ARHGAP36', 'ADAM33', 'FAM19A4']
	fbeta: 0.8707865168539326
12 out of 16:
	i5_i47_TRPC3_Ndnf_Car4
	 Pre-selected 1519 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  75%|████████████████████████████████████            | 12/16 [00:50<00:17,  4.41s/it]









    
    






	['NTNG1', 'EYA4']
	fbeta: 0.9064327485380117
13 out of 16:
	i6_i44_GPR149_Vip_Mybpc1
	 Pre-selected 670 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  81%|███████████████████████████████████████         | 13/16 [00:54<00:13,  4.44s/it]









    
    






	['FLT1', 'GPR149']
	fbeta: 0.7916666666666666
14 out of 16:
	i7_i31_CLMP_Ndnf_Cxcl14
	 Pre-selected 1599 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  88%|██████████████████████████████████████████      | 14/16 [00:59<00:09,  4.52s/it]









    
    






	['PAX6', 'TGFBR2']
	fbeta: 0.9009009009009009
15 out of 16:
	i8_i27_SNCG_Vip_Mybpc1
	 Pre-selected 2031 genes to feed into Random Forest.










    
    






running NSForest on all clusters:  94%|█████████████████████████████████████████████   | 15/16 [01:04<00:04,  4.80s/it]









    
    






	['SNCG', 'EDNRA']
	fbeta: 0.759493670886076
16 out of 16:
	i9_i22_TAC3_Vip_Mybpc1
	 Pre-selected 1477 genes to feed into Random Forest.










    
    






running NSForest on all clusters: 100%|████████████████████████████████████████████████| 16/16 [01:09<00:00,  4.36s/it]










    
    






	['MCTP2', 'VWC2L_IT1']
	fbeta: 0.7051282051282052
Saving supplementary table as...
../outputs_layer1/cluster_supplementary.csv
Saving markers table as...
../outputs_layer1/cluster_markers.csv
Saving results table as...
../outputs_layer1/cluster_results.csv
using median










    
    






Calculating medians (means) per cluster: 100%|████████████████████████████████████████| 16/16 [00:00<00:00, 356.51it/s]









    
    






Saving final results table as...
../outputs_layer1/cluster_results.csv
--- 69.85787081718445 seconds ---










    
    





























In [16]:



     


results



     






















    
    
Out[16]:











  
    

      
      clusterName
      clusterSize
      f_score
      PPV
      TN
      FP
      FN
      TP
      marker_count
      NSForest_markers
      binary_genes
      onTarget
    

  
  
    

      0
      e1_e299_SLC17A7_L5b_Cdh13
      299
      0.985003
      0.989726
      569
      3
      10
      289
      2
      [TESPA1, SATB2]
      [SLC17A7, TBR1, LINC00507, KIAA1211L, TESPA1, ...
      0.421602
    

    

      1
      g1_g48_GLI3_Astro_Gja1
      48
      0.950000
      1.000000
      823
      0
      10
      38
      1
      [LINC00498]
      [LINC00498, FGFR3, EMX2OS, SLC7A11, F3, RANBP3...
      1.000000
    

    

      2
      g2_g27_APBB1IP_Micro_Ctss
      27
      0.975610
      1.000000
      844
      0
      3
      24
      2
      [CSF2RA, ADAM28]
      [INPP5D, PLCG2, CSF2RA, ADAM28, FLI1, CD74, SY...
      1.000000
    

    

      3
      g3_g18_GPNMB_OPC_Pdgfra
      18
      0.862069
      1.000000
      853
      0
      8
      10
      2
      [GPNMB, OLIG2]
      [GPNMB, STK32A, COL20A1, OLIG2, KLRC3, B3GNT7,...
      1.000000
    

    

      4
      g4_g9_MOG_Oligo_Opalin
      9
      1.000000
      1.000000
      862
      0
      0
      9
      1
      [ST18]
      [ST18, MOBP, CNDP1, FOLH1, MOG, CD22, TF, CERC...
      1.000000
    

    

      5
      i10_i16_TSPAN12_Vip_Mybpc1
      16
      0.803571
      0.900000
      854
      1
      7
      9
      2
      [TSPAN12, CHRNB3]
      [TSPAN12, LINC01539, TMC5, CHRNB3, FAM46A, ANG...
      0.783762
    

    

      6
      i11_i6_EGF_Vip_Mybpc1
      6
      0.714286
      1.000000
      865
      0
      4
      2
      2
      [EGF, NRG3_AS1]
      [KCNJ2_AS1, EGF, ZSCAN23, TEKT1, STK11IP, NRG3...
      0.892726
    

    

      7
      i1_i90_COL5A2_Ndnf_Car4
      90
      0.887097
      1.000000
      781
      0
      35
      55
      2
      [COL5A2, FREM1]
      [NMBR, COL5A2, SST, NPNT, NDNF, FREM1, FAT1, G...
      0.556021
    

    

      8
      i2_i77_LHX6_Sst_Cbln4
      77
      0.816619
      0.838235
      783
      11
      20
      57
      1
      [LHX6]
      [LHX6, FLT3, TAC1, CALB1, TRBC2, RSPO3, GRIK3,...
      1.000000
    

    

      9
      i3_i56_BAGE2_Ndnf_Cxcl14
      56
      0.781250
      0.961538
      814
      1
      31
      25
      2
      [BAGE2, SYT10]
      [BAGE2, SCN5A, GREM2, SYT10, ARHGAP18, GRB14, ...
      0.602383
    

    

      10
      i4_i54_MC4R_Ndnf_Cxcl14
      54
      0.870787
      1.000000
      817
      0
      23
      31
      3
      [ARHGAP36, ADAM33, FAM19A4]
      [ARHGAP36, MC4R, COBLL1, LINC01435, ADAM33, FA...
      0.420466
    

    

      11
      i5_i47_TRPC3_Ndnf_Car4
      47
      0.906433
      1.000000
      824
      0
      16
      31
      2
      [NTNG1, EYA4]
      [SSTR2, KIRREL, NTNG1, TARID, EYA4, CA2, TACR1...
      0.380471
    

    

      12
      i6_i44_GPR149_Vip_Mybpc1
      44
      0.791667
      1.000000
      827
      0
      25
      19
      2
      [FLT1, GPR149]
      [PLCE1_AS1, FLT1, CXCL12, SLC22A3, GPR149, SHI...
      0.811541
    

    

      13
      i7_i31_CLMP_Ndnf_Cxcl14
      31
      0.900901
      1.000000
      840
      0
      11
      20
      2
      [PAX6, TGFBR2]
      [FGF10, KIAA1644, CLMP, PAX6, SP8, TGFBR2, WIF...
      0.547845
    

    

      14
      i8_i27_SNCG_Vip_Mybpc1
      27
      0.759494
      0.923077
      843
      1
      15
      12
      2
      [SNCG, EDNRA]
      [SNCG, MMRN2, EDNRA, FBN3, KCNK2, SCML4, SAMD5...
      1.000000
    

    

      15
      i9_i22_TAC3_Vip_Mybpc1
      22
      0.705128
      0.785714
      846
      3
      11
      11
      2
      [MCTP2, VWC2L_IT1]
      [BSPRY, MCTP2, OFD1P10Y, OFD1P15Y, OFD1P9Y, VW...
      0.883928
    

  































Evaluating input marker list¶


















In [17]:



     


# Getting dictionary of markers (format: {cluster: marker_list})
markers = pd.read_csv("../outputs_layer1/cluster_results.csv")
markers_dict = utils.prepare_markers(markers, "clusterName", "NSForest_markers")



     



















In [18]:



     


outputfilename_prefix = "marker_eval"
results = evaluating.DecisionTree(adata, cluster_header, medians_header, markers_dict, combinations = False, use_mean = False, 
                                  output_folder = output_folder, outputfilename_prefix = outputfilename_prefix)



     






















    
    






Preparing data...
--- 0.00398707389831543 seconds ---
Number of clusters to evaluate: 16
1 out of 16:
	e1_e299_SLC17A7_L5b_Cdh13
	marker genes to be evaluated: ['TESPA1', 'SATB2']
	['TESPA1', 'SATB2']
	f-beta:0.9850034083162916
	PPV:0.9897260273972602
2 out of 16:
	g1_g48_GLI3_Astro_Gja1
	marker genes to be evaluated: ['LINC00498']
	['LINC00498']
	f-beta:0.9500000000000001
	PPV:1.0
3 out of 16:
	g2_g27_APBB1IP_Micro_Ctss
	marker genes to be evaluated: ['CSF2RA', 'ADAM28']
	['CSF2RA', 'ADAM28']
	f-beta:0.9756097560975611
	PPV:1.0
4 out of 16:
	g3_g18_GPNMB_OPC_Pdgfra
	marker genes to be evaluated: ['GPNMB', 'OLIG2']
	['GPNMB', 'OLIG2']
	f-beta:0.8620689655172413
	PPV:1.0
5 out of 16:
	g4_g9_MOG_Oligo_Opalin
	marker genes to be evaluated: ['ST18']
	['ST18']
	f-beta:1.0
	PPV:1.0
6 out of 16:
	i10_i16_TSPAN12_Vip_Mybpc1
	marker genes to be evaluated: ['TSPAN12', 'CHRNB3']
	['TSPAN12', 'CHRNB3']
	f-beta:0.8035714285714286
	PPV:0.9
7 out of 16:
	i11_i6_EGF_Vip_Mybpc1
	marker genes to be evaluated: ['EGF', 'NRG3_AS1']
	['EGF', 'NRG3_AS1']
	f-beta:0.7142857142857143
	PPV:1.0
8 out of 16:
	i1_i90_COL5A2_Ndnf_Car4
	marker genes to be evaluated: ['COL5A2', 'FREM1']
	['COL5A2', 'FREM1']
	f-beta:0.8870967741935484
	PPV:1.0
9 out of 16:
	i2_i77_LHX6_Sst_Cbln4
	marker genes to be evaluated: ['LHX6']
	['LHX6']
	f-beta:0.8166189111747851
	PPV:0.8382352941176471
10 out of 16:
	i3_i56_BAGE2_Ndnf_Cxcl14
	marker genes to be evaluated: ['BAGE2', 'SYT10']
	['BAGE2', 'SYT10']
	f-beta:0.78125
	PPV:0.9615384615384616
11 out of 16:
	i4_i54_MC4R_Ndnf_Cxcl14
	marker genes to be evaluated: ['ARHGAP36', 'ADAM33', 'FAM19A4']
	['ARHGAP36', 'ADAM33', 'FAM19A4']
	f-beta:0.8707865168539326
	PPV:1.0
12 out of 16:
	i5_i47_TRPC3_Ndnf_Car4
	marker genes to be evaluated: ['NTNG1', 'EYA4']
	['NTNG1', 'EYA4']
	f-beta:0.9064327485380117
	PPV:1.0
13 out of 16:
	i6_i44_GPR149_Vip_Mybpc1
	marker genes to be evaluated: ['FLT1', 'GPR149']
	['FLT1', 'GPR149']
	f-beta:0.7916666666666666
	PPV:1.0
14 out of 16:
	i7_i31_CLMP_Ndnf_Cxcl14
	marker genes to be evaluated: ['PAX6', 'TGFBR2']
	['PAX6', 'TGFBR2']
	f-beta:0.9009009009009009
	PPV:1.0
15 out of 16:
	i8_i27_SNCG_Vip_Mybpc1
	marker genes to be evaluated: ['SNCG', 'EDNRA']
	['SNCG', 'EDNRA']
	f-beta:0.759493670886076
	PPV:0.9230769230769231
16 out of 16:
	i9_i22_TAC3_Vip_Mybpc1
	marker genes to be evaluated: ['MCTP2', 'VWC2L_IT1']
	['MCTP2', 'VWC2L_IT1']
	f-beta:0.7051282051282052
	PPV:0.7857142857142857
using median










    
    






Calculating medians (means) per cluster: 100%|████████████████████████████████████████| 16/16 [00:00<00:00, 485.67it/s]









    
    






Saving final results table as...
../outputs_layer1/marker_eval_results.csv
--- 0.5076773166656494 seconds ---










    
    





























In [19]:



     


results



     






















    
    
Out[19]:











  
    

      
      clusterName
      clusterSize
      f_score
      PPV
      TN
      FP
      FN
      TP
      marker_count
      markers
      onTarget
    

  
  
    

      0
      e1_e299_SLC17A7_L5b_Cdh13
      299
      0.985003
      0.989726
      569
      3
      10
      289
      2
      [TESPA1, SATB2]
      0.421602
    

    

      1
      g1_g48_GLI3_Astro_Gja1
      48
      0.950000
      1.000000
      823
      0
      10
      38
      1
      [LINC00498]
      1.000000
    

    

      2
      g2_g27_APBB1IP_Micro_Ctss
      27
      0.975610
      1.000000
      844
      0
      3
      24
      2
      [CSF2RA, ADAM28]
      1.000000
    

    

      3
      g3_g18_GPNMB_OPC_Pdgfra
      18
      0.862069
      1.000000
      853
      0
      8
      10
      2
      [GPNMB, OLIG2]
      1.000000
    

    

      4
      g4_g9_MOG_Oligo_Opalin
      9
      1.000000
      1.000000
      862
      0
      0
      9
      1
      [ST18]
      1.000000
    

    

      5
      i10_i16_TSPAN12_Vip_Mybpc1
      16
      0.803571
      0.900000
      854
      1
      7
      9
      2
      [TSPAN12, CHRNB3]
      0.783762
    

    

      6
      i11_i6_EGF_Vip_Mybpc1
      6
      0.714286
      1.000000
      865
      0
      4
      2
      2
      [EGF, NRG3_AS1]
      0.892726
    

    

      7
      i1_i90_COL5A2_Ndnf_Car4
      90
      0.887097
      1.000000
      781
      0
      35
      55
      2
      [COL5A2, FREM1]
      0.556021
    

    

      8
      i2_i77_LHX6_Sst_Cbln4
      77
      0.816619
      0.838235
      783
      11
      20
      57
      1
      [LHX6]
      1.000000
    

    

      9
      i3_i56_BAGE2_Ndnf_Cxcl14
      56
      0.781250
      0.961538
      814
      1
      31
      25
      2
      [BAGE2, SYT10]
      0.602383
    

    

      10
      i4_i54_MC4R_Ndnf_Cxcl14
      54
      0.870787
      1.000000
      817
      0
      23
      31
      3
      [ARHGAP36, ADAM33, FAM19A4]
      0.420466
    

    

      11
      i5_i47_TRPC3_Ndnf_Car4
      47
      0.906433
      1.000000
      824
      0
      16
      31
      2
      [NTNG1, EYA4]
      0.380471
    

    

      12
      i6_i44_GPR149_Vip_Mybpc1
      44
      0.791667
      1.000000
      827
      0
      25
      19
      2
      [FLT1, GPR149]
      0.811541
    

    

      13
      i7_i31_CLMP_Ndnf_Cxcl14
      31
      0.900901
      1.000000
      840
      0
      11
      20
      2
      [PAX6, TGFBR2]
      0.547845
    

    

      14
      i8_i27_SNCG_Vip_Mybpc1
      27
      0.759494
      0.923077
      843
      1
      15
      12
      2
      [SNCG, EDNRA]
      1.000000
    

    

      15
      i9_i22_TAC3_Vip_Mybpc1
      22
      0.705128
      0.785714
      846
      3
      11
      11
      2
      [MCTP2, VWC2L_IT1]
      0.883928
    

  































Plotting¶


















In [20]:



     


ns.pl.boxplot_fscore(results, False, output_folder, outputfilename_prefix)



     






















    
    






        
        









    
    







                            
                    
























In [21]:



     


ns.pl.boxplot_ppv(results, False, output_folder, outputfilename_prefix)



     






















    
    







                            
                    
























In [22]:



     


ns.pl.boxplot_ontarget(results, False, output_folder, outputfilename_prefix)



     






















    
    







                            
                    
























In [23]:



     


ns.pl.scatter_w_clusterSize_fscore(results, False, output_folder, outputfilename_prefix)



     






















    
    







                            
                    
























In [24]:



     


ns.pl.scatter_w_clusterSize_ppv(results, False, output_folder, outputfilename_prefix)



     






















    
    







                            
                    
























In [25]:



     


ns.pl.scatter_w_clusterSize_ontarget(results, False, output_folder, outputfilename_prefix)



     






















    
    







                            
                    
























In [26]:



     


# If you want to specify dendrogram order
dendrogram = list(adata.uns["dendrogram_" + cluster_header]["categories_ordered"])
if dendrogram: # https://stackoverflow.com/questions/23482668/sorting-by-a-custom-list-in-pandas
    results["clusterName"] = results["clusterName"].astype("category")
    results["clusterName"] = results["clusterName"].cat.set_categories(dendrogram)
    results = results.sort_values("clusterName")



     



















In [27]:



     


if "markers" in list(results.columns): 
    markers_dict = dict(zip(results["clusterName"], results["markers"]))
elif "NSForest_markers" in list(results.columns): 
    markers_dict = dict(zip(results["clusterName"], results["NSForest_markers"]))



     



















In [31]:



     


ns.pl.dotplot(adata, markers_dict, cluster_header, dendrogram, True, output_folder, outputfilename_prefix)



     






















    
    






Saving...
 marker_eval.png
WARNING: saving figure to file figures\dotplot_marker_eval.png










    
    






C:\Users\bpeng\AppData\Local\anaconda3\Lib\site-packages\scanpy\plotting\_dotplot.py:747: UserWarning:

No data for colormapping provided via 'c'. Parameters 'cmap', 'norm' will be ignored











    
    































In [32]:



     


ns.pl.stackedviolin(adata, markers_dict, cluster_header, dendrogram, True, output_folder, outputfilename_prefix)



     






















    
    






Saving...
 marker_eval.png
WARNING: saving figure to file figures\stacked_violin_marker_eval.png










    
    































In [33]:



     


ns.pl.matrixplot(adata, markers_dict, cluster_header, dendrogram, True, output_folder, outputfilename_prefix)



     






















    
    






Saving...
 marker_eval.png
WARNING: saving figure to file figures\matrixplot_marker_eval.png